Development and psychometric testing of a rehabilitation specific Comorbidity score

Background and purpose: It is extremely important to consider the comorbidity and multimorbidity of patients - especially for rehabilitation. There are some internationally established indexes for measuring multimorbidity. However, such an instrument designed explicitly for rehabilitation has not yet been published in the German-speaking world. In this article, a generic, rehabilitation-specific instrument to assess comorbidity and multimorbidity with a focus on functionality (KoMo score) is introduced and evaluated as to the method's quality criteria. Methods: The KoMo score was developed using established comorbidity indexes and is available in a physician and a patient version. The concomitant diseases surveyed can be aggregated to a total number of diseases score, a degree of severity score, a activity limitations score, and a total comorbidity score. The examination of the quality criteria of the physician version includes analyses of practicability, construct validity, and concurrent and predictive validity. The data were collected from N = 998 inpatient rehab subjects at nine clinics and N = 654 outpatient rehab subjects at seven rehabilitation centers, all of whom were being treated for musculoskeletal diseases. Results: The practicability of the KoMo score is shown by the low amount of missing values and a consistent approach. Correlations with age and baseline health status are an indication of the construct validity of the instrument, while the concurrent criterion validity is supported by anticipated variations between outpatient and inpatient rehab subjects. Demonstrating the predictive relevance of multimorbidity both in terms of baseline health status and of the outcome of rehabilitation gives evidence for the predictive validity of the score. Conclusion: These data are an indication of the practicability and validity of the rehabilitation KoMo score. Future areas for research include the further testing of the score, for example with respect to inter-rater reliability and test-retest-reliability, as well as assessing the patient version of the instrument developed in parallel with the physician version.